Report Contrasting ‘‘Fish’’ Diversity Dynamics between Marine and Freshwater Environments Highlights d Phylogenetic diversities of fish groups are compared with models of diversity dynamics d Marine diversity fits an equilibrium model, while freshwater diversity is in expansion d Greater freshwater competition, isolation, and perturbations account for this pattern d Fish diversity dynamics confirm previous models designed for the entire biosphere Authors Guillaume Guinot, Lionel Cavin Correspondence [email protected]In Brief Guinot and Cavin demonstrate that deep- time phylogenetic ‘‘fish’’ diversity dynamics fit an equilibrium model in marine environments and an expansion model in freshwater environments. ‘‘Fish’’ diversity dynamics support former empiric models designed for the whole biosphere, with a maximum carrying capacity in marine environments. Guinot & Cavin, 2015, Current Biology 25, 2314–2318 August 31, 2015 ª2015 Elsevier Ltd All rights reserved http://dx.doi.org/10.1016/j.cub.2015.07.033
13
Embed
Contrasting ``Fish'' Diversity Dynamics between Marine and ... · PDF fileReport Contrasting ‘‘Fish’’ Diversity Dynamics between Marine and Freshwater...
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Report
Contrasting ‘‘Fish’’ Divers
ity Dynamics betweenMarine and Freshwater Environments
Highlights
d Phylogenetic diversities of fish groups are compared with
models of diversity dynamics
d Marine diversity fits an equilibrium model, while freshwater
diversity is in expansion
d Greater freshwater competition, isolation, and perturbations
account for this pattern
d Fish diversity dynamics confirm previous models designed
for the entire biosphere
Guinot & Cavin, 2015, Current Biology 25, 2314–2318August 31, 2015 ª2015 Elsevier Ltd All rights reservedhttp://dx.doi.org/10.1016/j.cub.2015.07.033
Contrasting ‘‘Fish’’ Diversity Dynamicsbetween Marine and Freshwater EnvironmentsGuillaume Guinot1,2,* and Lionel Cavin11Natural HistoryMuseumof Geneva, Department of Geology and Palaeontology, Route deMalagnou 1, CP 6434, 1211Geneva 6, Switzerland2Present address: Institut des Sciences de l’Evolution, UMR-CNRS 5554, CC064, Universite de Montpellier, Place Eugene Bataillon, 34095
Two theoretical models have been proposed todescribe long-term dynamics of diversification: theequilibrium model considers the Earth as a closedsystem with a fixed maximum biological carryingcapacity, whereas the expansion model hypothe-sizes a continuously increasing diversification oflife. Based on the analysis of the fossil record of allorganisms, Benton [1] suggested contrastingmodelsof diversity dynamics between marine and con-tinental realms. Diversity in marine environments ischaracterized by phases of rapid diversification fol-lowed by plateaux, i.e., an equilibrium model [2–4]directly derived from insular biogeography theories[5, 6], whereas diversity in continental environmentsis characterized by exponential growth. Previousstudies that aimed at testing these models withempirical data were based on datasets extracteddirectly from the reading of the vagaries of the rawfossil record, without correcting for common fossilrecord biases (preservation and sampling). Althoughcorrection of datasets for the incompleteness ofthe fossil record is now commonly performed for ad-dressing long-term biodiversity variations [7, 8], onlya few attempts [9] have been made to produce diver-sity curves corrected by phylogenetic data fromextant and extinct taxa. Here we show that phyloge-netically corrected diversity curves for ‘‘fish’’ (actino-pterygians and elasmobranchs) during the last 200million years fit an equilibrium model in the marinerealm and an expansion model in the freshwaterrealm. These findings demonstrate that the rate ofdiversification has decreased for marine fish overthe Cenozoic but is in sharp expansion for freshwaterfish.
RESULTS AND DISCUSSION
Here we test the fit betweenmathematical models and corrected
diversity curves for two aquatic vertebrate groups (elasmo-
branchs and ray-finned fishes) based on phylogenetic diversities
including both fossil and living taxa. Corrected diversity curves
were computed by adding to observed temporal ranges of
taxa (read directly from the fossil record) the ghost lineages to
accommodate first appearance age of taxa with their corre-
sponding phylogenetic relationships. Both fish clades together
account for more than half of total vertebrate diversity and
constitute about 10% (9.7%) of aquatic animal diversity as well
as almost 83% of aquatic vertebrate diversity (Table 1). In addi-
tion, the evolutionary history of ray-finned fishes encompasses
three of the largest diversifications among jawed vertebrates
[15], including the biggest (percomorphs). In view of these char-
acteristics, we consider that the diversity trajectories of these
groups are good proxies for assessing global diversity patterns
in the marine and freshwater realms.
Fish diversities considered here span the Late Triassic to
Recent interval for elasmobranchs and the Late Jurassic to
Recent interval for actinopterygians, at family level. In a previous
study [16], we provided a range of computed diversity values
according to the various phylogenies considered and their cor-
responding congruence with the fossil record, indicating that
genuine diversity values should lie within this range. Conse-
quently, the median diversity value was selected here for each
time bin in order to sum up all hypotheses in one curve. The total
actinopterygian dataset is divided into three subsets: fully ma-
rine, fully freshwater, and mixed-environment taxa (see Data
S1). The latter subset encompasses clades that include either
taxa from both freshwater and marine environments or euryha-
line taxa (salt-tolerant and diadromous fishes). The marine
actinopterygian subset and elasmobranch data were merged
in order to provide a ‘‘total marine fish’’ dataset. Observed and
computed data were compared with mathematical models that
are commonly proposed to represent the main theoretical diver-
sification dynamics of biological organisms. These include the
additive, expansionist, and equilibrium models, represented
mathematically by the linear, exponential, and logistic functions,
respectively. In addition, the quadratic polynomial function (e.g.,
polynomial of degree 2) was included as an alternative represen-
tation (in the case of a negative discriminant) of the expansionist
theoretical model of diversification. Model selection was per-
formed using the Akaike information criterion with correction
for finite/small sample sizes (AICc) (see Supplemental Experi-
mental Procedures).
Fits of the various models of diversity dynamics to the main
‘‘fish’’ (here, actinopterygians and elasmobranchs) diversity
datasets considered here are provided in Table 2 (see Table
S1 for detailed results). With the exception of the freshwater
2314 Current Biology 25, 2314–2318, August 31, 2015 ª2015 Elsevier Ltd All rights reserved
See Table S1 for results on complete datasets. Scores indicating best model fit (lowest AICc, DAICc = 0) are indicated with asterisks (*). Logis, logistic;
Lin, linear; Exp, exponential; Poly, second-degree polynomial (quadratic polynomial). ‘‘�Extant’’ indicates that the value corresponding to today’s di-
versity was removed.
2316 Current Biology 25, 2314–2318, August 31, 2015 ª2015 Elsevier Ltd All rights reserved
primary producers and zooplankton whereas freshwater ecosys-
tems also largely rely on detrital foodwebs. Hence, marine diver-
sity depends on the fluctuations of phyto- and zooplankton,
which are themselves linked with environmental forcing and
therefore more prone to extinctions through time, whereas con-
tinental food webs sustain less perturbation. Tectonics, and to
a greater extent the evolution of Earth’s geographical and envi-
ronmental configuration, are another factor that may explain
the contrasting deep-time evolution of the freshwater andmarine
‘‘fish’’ diversities. The Mesozoic-Cenozoic interval is char-
acterized mainly by the breakup of Pangaea, which provided
increasing ecological niches in both marine and continental eco-
systems [16]. In the marine realm, it has been shown that periods
of high sea levels coupledwith warm global temperatures (Upper
Cretaceous, Paleocene-Eocene) are linked with major diversifi-
cation events within ‘‘fishes’’ [16] and more broadly vertebrates
[15], along with habitat complexification related to the settlement
of modern reef ecosystems (Paleocene-Eocene). Similarly, di-
versity in continental ‘‘fish’’ faunas seems to have been positively
affected by high temperatures and sea level variations, but also
by periods of heterogeneous global heat distribution, including
monsoonal activities in the Lower Cretaceous [16]. Although
post-Eocene marine geography and eustasy have undergone
relatively few important perturbations until the present day (in
comparisonwith pre-Oligocene times), this period encompasses
marked climatic fluctuations (glaciations, temperature gradients)
and major orogenesis and rifting events in the continental realm
that deeply modified regional climatic settings and river net-
works. These still-ongoing processes shaped new continental
hydrographic systems, increased the complexity of continental
aquatic environments and atmospheric circulations, and are
possible factors in the higher carrying capacity (if any) of fresh-
water ecosystems in comparison to the marine realm.
Our survey of ‘‘fish’’ diversity dynamics covers a short portion
of the complete history of the metazoan evolution, but it covers
most of the Modern Fauna time interval as defined by Sepkoski
[27], which is characterized by the expansion of chondrichthyan
and osteichthyan ‘‘fishes,’’ among others. Based on this ascer-
tainment and the large proportion of aquatic vertebrate diversity
represented by ‘‘fish,’’ the distinctions found here between the
models for marine and freshwater realms are regarded as reflect-
ing global features associated with these peculiar environments,
which impact how life diversifies.
SUPPLEMENTAL INFORMATION
Supplemental Information includes two tables, Supplemental Experimental
Procedures, and one dataset and can be found with this article online at
http://dx.doi.org/10.1016/j.cub.2015.07.033.
ACKNOWLEDGMENTS
The authors wish to thank J. Claude for helpful comments on methodology as
well as the editor and two anonymous reviewers for their comments on an
earlier version of this paper. This paper is a contribution to the project ‘‘Fish
Response to Long-Term Global Changes’’ supported by the Swiss National
Science Foundation (200021-140827).
Phyl
ogen
etic
div
ersi
ty (N
umbe
r of f
amili
es)
Phyl
ogen
etic
div
ersi
ty (N
umbe
r of f
amili
es)
Figure 1. Curves of Best-Fitting Models
Plotted over Phylogenetic Family-Level
Diversity through Geological Times
Gray dots represent values of corrected diversity
(phylogenetic diversity) per geological stage. All
marine datasets (blue) fit a logistic model (repre-
senting the equilibrium model of evolution),
whereas the freshwater ray-finned data fit a
quadratic polynomial function (representing the
expansion model of evolution). Curves are scaled
to zero for graphical purpose. Jur., Jurassic;
Paleog., Paleogene; Ng., Neogene.
Current Biology 25, 2314–2318, August 31, 2015 ª2015 Elsevier Ltd All rights reserved 2317
Scores indicating best model fit (lowest AICc, Δ AICc = 0) are indicated with asterisks (*). Logis, logistic; Lin, linnear; Exp, exponential; Poly, second-degree polynomial
(quadratic polynomial). “− Extant” indicates that the value corresponding to today’s diversity was removed.
3
Supplemental Table S2. Fit of the genus-level diversity dataset for elasmobranchs to four theoretical
models of diversification.
Elasmobranchs (genera)
AICc ΔAICc wAICc
Computed diversity
Logis 471.537337* 0* 0.99809739*
Lin 516.1433 44.6059626 2.06E-10
Exp 527.707368 56.1700309 6.34E-13
Poly 484.062582 12.5252445 0.00190261
Observed diversity
Logis 435.372023* 0* 0.80975498*
Lin 438.867331 3.49530749 0.14104485
Exp 448.83328 13.461257 0.00096667
Poly 441.013379 5.64135596 0.0482335
Computed diversity (-Extant)
Logis 463.127132* 0* 0.99701072*
Lin 506.902861 43.7757296 3.11E-10
Exp 518.264684 55.137552 1.06E-12
Poly 474.746588 11.6194563 0.00298928
Observed diversity (-Extant)
Logis 382.591979* 0* 0.99981523*
Lin 400.444799 17.8528196 0.00013281
Exp 420.982944 38.3909648 4.61E-09
Poly 402.322009 19.7300295 5.20E-05
See Supplemental Table S1 for details.
2. Supplemental experimental procedures
2.1. Data sets
The ‘fish’ diversity data used here were taken from a recent study [1] that provided
elasmobranch and actinopterygian supertrees including both extant and extinct families along with
the fossil record (ages of first and last occurrence) of corresponding terminal taxa. For both groups,
four competing tree topologies representing alternative phylogenetic hypotheses were plotted
against the fossil record of terminal taxa. This resulted in the addition of artificial stratigraphic range
(ghost range) to the observed fossil record of a taxon to fit the first appearance date of its sister
4
taxon (two sister taxa must have the same age of first occurrence). Comparing stratigraphic ranges of
taxa and corresponding phylogenetic relationships requires dealing with uncertainties related to
each datasets, namely the range age of first occurrence (stratigraphy) and polytomies (phylogeny).
Consequently, the method used in Guinot & Cavin [1] followed that of Boyd et al. [2] for measuring
congruence scores of the fit of stratigraphic data to phylogenies. This resolves polytomies in two
ways: one Chronological where the original polytomous clade is resolved in a pectinate arrangement,
placing taxa with the oldest age of first occurrence at the base of the clade and one Reverse
Chronological polytomy resolution where taxa with the youngest age of first occurrence are placed at
the base of the resolved clade. Uncertainties with the age of first occurrence were considered in
randomly picking an age within the age range of first occurrence of each taxon, using 1 000 000
replicates. Hence, for each of the four phylogenetic hypotheses considered, range values of
computed first appearance ages are provided for the Chronological method and Reverse